Date: Sun, 21 Apr 2019 05:16:48 +0000 From: bugzilla-noreply@freebsd.org To: ppc@FreeBSD.org Subject: [Bug 233863] r345425 on PowerMac G5 may require kern.smp.disabled=1 and must set usefdt=1 which causes net interface reorder Message-ID: <bug-233863-21-WaCWPuTYwH@https.bugs.freebsd.org/bugzilla/> In-Reply-To: <bug-233863-21@https.bugs.freebsd.org/bugzilla/>
index | next in thread | previous in thread | raw e-mail
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=233863 --- Comment #23 from Mark Millard <marklmi26-fbsd@yahoo.com> --- (In reply to Mark Millard from comment #20) The original rejection by a debug build that lead to the discovery of out of bounds access was tied to the original conversion to fdt code truncating the translation property via: if (proplen > 1024) { proplen = 1024; } in add_node_to_fdt in stand/powerpc/ofw/ofwfdt.c . This changed a 1040==208*5 total to a 1024==256*4 total. (1024 is not a multiple of 5.) So the problem goes away when the truncation logic is removed, which is part of what my patches do. Still, the truncation did expose some coding problems in the translation map extraction, such as out of bounds access for such a truncated case. But it would take some forced bad property size to see the problem again if mmu_oea64.c is not patched. The change to the KASSERT in my patch may well be inappropriate, given the above context that is now known. Having an incomplete set of translations does not seem like an appropriate thing: so the truncation to 1024 needs to be avoided. -- You are receiving this mail because: You are the assignee for the bug.home | help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-233863-21-WaCWPuTYwH>